Cortical AAV-CNTF Gene Therapy Combined with Intraspinal ...downloads.hindawi.com/journals/np/2018/9828725.pdf · adult CNS. Here we tested whether CNTF gene therapy targeting corticospinal

Research ArticleCortical AAV-CNTF Gene Therapy Combined with IntraspinalMesenchymal Precursor Cell Transplantation PromotesFunctional and Morphological Outcomes after Spinal CordInjury in Adult Rats

Stuart I. Hodgetts ,1,2 Jun Han Yoon,1 Alysia Fogliani,1 Emmanuel A. Akinpelu,1

Danii Baron-Heeris,1 Imke G. J. Houwers,1 Lachlan P. G. Wheeler,1 Bernadette T. Majda,3

Sreya Santhakumar,1,2 Sarah J. Lovett,1 Emma Duce,1 Margaret A. Pollett,1

Tylie M. Wiseman,1 Brooke Fehily,1 and Alan R. Harvey 1,2

1School of Human Sciences, The University of Western Australia (UWA), Perth, WA 6009, Australia2Perron Institute for Neurological and Translational Science, Nedlands, WA 6009, Australia3University of Notre Dame Australia, Fremantle, WA 6959, Australia

Correspondence should be addressed to Stuart I. Hodgetts; [email protected]

Received 26 January 2018; Revised 18 April 2018; Accepted 21 May 2018; Published 6 August 2018

Academic Editor: Michele Fornaro

Copyright © 2018 Stuart I. Hodgetts et al. This is an open access article distributed under the Creative Commons AttributionLicense, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work isproperly cited.

Ciliary neurotrophic factor (CNTF) promotes survival and enhances long-distance regeneration of injured axons in parts of theadult CNS. Here we tested whether CNTF gene therapy targeting corticospinal neurons (CSN) in motor-related regions of thecerebral cortex promotes plasticity and regrowth of axons projecting into the female adult F344 rat spinal cord after moderatethoracic (T10) contusion injury (SCI). Cortical neurons were transduced with a bicistronic adeno-associated viral vector (AAV1)expressing a secretory form of CNTF coupled to mCHERRY (AAV-CNTFmCherry) or with control AAV only (AAV-GFP) twoweeks prior to SCI. In some animals, viable or nonviable F344 rat mesenchymal precursor cells (rMPCs) were injected into thelesion site two weeks after SCI to modulate the inhibitory environment. Treatment with AAV-CNTFmCherry, as well as withAAV-CNTFmCherry combined with rMPCs, yielded functional improvements over AAV-GFP alone, as assessed by open-fieldand Ladderwalk analyses. Cyst size was significantly reduced in the AAV-CNTFmCherry plus viable rMPC treatment group.Cortical injections of biotinylated dextran amine (BDA) revealed more BDA-stained axons rostral and alongside cysts in theAAV-CNTFmCherry versus AAV-GFP groups. After AAV-CNTFmCherry treatments, many sprouting mCherry-immunopositiveaxons were seen rostral to the SCI, and axons were also occasionally found caudal to the injury site. These data suggest thatCNTF has the potential to enhance corticospinal repair by transducing parent CNS populations.

1. Introduction

Most spinal cord injury (SCI) results from contusion ratherthan transection injuries, and cervical injuries (~60–70% ofall SCI) produce greater deficits and threaten more criticalsurvival systems than thoracic/lumbar SCI. The corticospinaltract (CST) is important in the control of voluntary skilledmovements, especially of distal limbs. Human CST projec-tions are not completely homologous to the descending

CST in rodents (which in these species projects mainly tothe forelimbs and is mainly located in the dorsal rather thanlateral columns) [1], nonetheless, because of the importanceof CST projections in fine manipulatory motor control, thispathway has been a focus of many experimental repair strat-egies aimed at restoring function following SCI. Most studiesin rodents have attempted this by the delivery of purified neu-rotrophic growth factors and/or by cell transplantation usingdonor cells engineered to overexpress the growth factors, the

HindawiNeural PlasticityVolume 2018, Article ID 9828725, 15 pageshttps://doi.org/10.1155/2018/9828725

factors usually applied to the injury site itself (see [2]). In suchstudies, functional improvements usually reflect sproutingand some plasticity in collateral and/or intraspinal pathways([3, 4], c.f. [5]), rather than axonal regeneration per se.

Our gene therapy approach targets corticospinal neurons(CSN) and is aimed at enhancing axonal plasticity and induc-ing regeneration of CST axons, leading to behaviouralimprovements after SCI. Gene therapy involves the transduc-tion of neurons in the sensorimotor cortex using a bicistronicadeno-associated viral vector (AAV) that encodes andexpresses a secretory form of ciliary neurotrophic factor(CNTF) coupled to mCHERRY (AAV-CNTFmCherry). Injec-tion of AAV-CNTFmCherry into the cortical regions of thebrain that project onto the output pathways of the CSTallows expression of CNTF in neurons, including CSN, atthe time of SCI.

CNTF has been selected because it is known to promotethe survival of injured CSN [6], and in other systems it pro-motes long-distance regeneration of injured adult CNS axons[7–10], with at least some functional outcomes [11]. CNTF isa neuropoietic member of the interleukin 6 (IL-6) cytokinefamily, expressed primarily in glial cells of the nervous sys-tem [12–15]. Survival effects of CNTF have been demon-strated on motor, retinal ganglion, cortical and hippocampal,red nucleus, and striatal and thalamic neuron populations[14–16] (see also [17]). CNTF has demonstrated efficacy inlimiting neuronal injury in several experimental disease andinjury paradigms, and has been clinically evaluated as a treat-ment for motor neuron-specific amyotrophic lateral sclerosis(ALS) [18, 19]. CNTF protects corticospinal neurons in thesensorimotor cortex after intracortical axotomy [6], and theseneurons, at least in murine neonates, are known to expressCNTF receptor α [20]. Functional improvement andenhanced remyelination has also been reported after intrasp-inal transplantation of oligodendrocyte precursor cellsexpressing CNTF [21]. However, as we have argued [2], whilemost SCI studies have targeted the injured cord itself for ther-apy, in other systems—such as the visual system—targetingthe injured neurons themselves yields excellent functionaloutcomes [11, 22]. Most importantly there is now clear evi-dence in human postmortem material that there is long-termsurvival of CSN after SCI [23], making strategies that targetthese neurons of genuine clinical relevance, potentially in bothacute and chronic circumstances.

In an initial study, we obtained preliminary evidence thatAAV2.1-CNTFmCherry transduced large numbers of neuronsin the sensorimotor cortex containing CSNs projecting intothe spinal cord, and after moderate thoracic T10 contusionSCI there was clear and consistent sprouting of mCherry-labelled CST axons at, and rostral to, the lesion site [2]. Inthe present report, cortical gene therapy has been also com-binedwith the transplantationofmesenchymal precursor cells(MPCs) into the injured thoracic cord, a method shown by usand others consistently to limit tissue loss and promotemorphological sparing as well as functional improvementfollowing SCI [24, 25]. The rationale is that a healthierlocal environment at the cell-transplant injury site providesa better terrain for plasticity and regeneration of CST axonsafter targeted CNTF expression in the motor cortex. MPC

treatment may be especially needed for contusion injurieswhich tend to be more unstable and prone to cystformation—this is vital when treating cervical injuries inhumans due to ongoing cavity formation (syringomyelia)and consequent progressive, sometimes catastrophic, loss offunction [26].

2. Materials and Methods

2.1. Animals and Experimental Design. Adult female Fischer(F344) rats (age 10–12wk, 120–150 g; Animal Resource Cen-tre, Western Australia) were used in experimental proce-dures conforming to National Health and Medical ResearchCouncil Guidelines (Australia) and approved by the Univer-sity of Western Australia Animal Ethics Committee. A totalof 43 rats was used, distributed between 4 experimentalSCI groups as follows; SCI+ control AAV-GFP (n = 11),SCI+AAV-CNTFmCherry (n = 16), SCI+AAV-CNTFmCherry +nonviable rMPCs (n = 6), and SCI+AAV-CNTFmCherry +viable rMPCs (n = 11). Previous experiments conducted inour lab used SCI + either nonviable or viable rMPC cell trans-plantation only [24, 25], and to satisfy NHMRC guidelines toaddress animal welfare and minimise animal use, thesegroups were not repeated for the present study. Viral AAVtransduction was performed 2 weeks prior to SCI, and2 weeks prior to experimental endpoint, biotinylateddextran amine BDA conjugated to horseradish peroxidase(Thermo Fisher Scientific) was spaced and injected in vir-tually identical positions as the earlier AAV injections inorder to cover the same hindlimb projection fields in thesensorimotor cortex. Endpoint was reached and animalsculled at day 56 after SCI.

2.1.1. Viral Vectors. The bicistronic adeno-associated viralvector (AAV) encoding and expressing a secretory form ofCNTF coupled to mCHERRY (AAV-CNTFmCherry) orvector-alone control linked to green fluorescent protein(AAV-GFP) were made by Vector Biolabs, USA. AAVvectors consisted of an AAV2 DNA backbone in an AAV1capsid. This particular serotype has been shown to provideexcellent transduction of cortical neurons [27–29]. Trans-gene expression was driven by a shortened CAG2 promoterbased on the typical cytomegalovirus/chicken beta-actin(CAG) promoter. For bicistronic vectors, the transgene(CNTF) and reporter (mCherry) were linked by a 2A viralpeptide sequence, which causes a “translational skip” andresults in a 1 : 1 expression of transgene and reporter pro-teins. The CNTF transgene is a mouse CNTF gene precededby the secretory signal sequence from mouse pre-pro-NGF,to allow for local secretion of CNTF [30] (a gift from Prof.M. Sendtner, University of Wurzburg, Germany).

2.1.2. Viral Transduction of Hindlimb Sensorimotor Cortex.AAV-CNTFmCherry or AAV-GFP transduction of CSN inthe cortical regions that project axons to the level of the spi-nal cord that controls the hindlimbs [2, 31] was performed 2weeks prior to SCI via 4× 0.5μl injections of the respectivevirus (~4× 1013 genomic copies/ml) using a nanojet deviceand a digital stereotaxic frame (Kopf) for accurate Bregma/

2 Neural Plasticity

Lambda coordinate positioning. Injection of AAV 2 weeksprior to SCI allowed time for onset of transgene expressionand production of CNTF.

2.2. Spinal Cord Injury (SCI). Rats were anesthetized with1.5% (v/v) halothane (Rhone-Poulenc Chemicals Pty. Ltd.,Australia) combined with nitrous oxide (60%) and oxygen(38.5%). Amacin ophthalmic eye ointment was applied beforerats were placed on a heating pad (37°C). Partial laminectomyat vertebral level T9-T10 exposed the SC underneath withoutdisrupting the dura [24, 25]. Using an Infinite Horizonimpactor device, a 200 kDyne contusion injury was inducedat the exposed spinal cord. Postsurgery care, analgesics, food,and housing were as previously described [24, 25]. Briefly,rats were treated with Benacillin (0.02ml/100 g body wt.,300U/ml, i.m.) and painkiller Buprenorphine (Temgesic,0.01ml/100 g, 300U/ml, ip) for 5 days.

2.3. Donor rMPCs and Transplantation. Commerciallyavailable rMPCs isolated from Fischer F344 rats (CyagenBiosciences Inc., number RAFMX-01201) were routinelymaintained in Mesenchymal Stem Cell Growth Medium(number GUXMX-90011) prior to use in transplantationexperiments, or for those used for routine phenotypic charac-terization and differentiation [24, 25] maintained for at least24–48 hr, in order to determine any neuronal phenotype/marker expression using the panel of antibodies describedin Section 2.7.2. For transplantation experiments, MPCs(no higher than passage number 5) were washed andresuspended in PBS. Nonviable MPCs were prepared bymultiple (×3) freeze/thawing steps between −80°C and 37°Cand confirming loss of viability using trypan blue stainingunder microscopy. At day 14 after SCI, 6× 105 cells in atotal of 4μl were injected in a single injection directly intothe lesion site (rostrocaudally at 1mm depth) through afinely drawn (80μm tip) glass pipette connected to a10μl Hamilton syringe and driven by a Harvard Pumpat 0.5μl/min (total duration is 8min). The pipette wasleft in place to prevent cell leakage for 1min beforewithdrawal [24, 25].

2.4. BDA Injection into Sensorimotor Cortex that Projects toLow Thoracic/Lumbar Spinal Cord. Injections of biotinylateddextran amine were spaced in virtually identical positions asthe earlier AAV-CNTFmCherry or AAV-GFP transductioninjections in order to cover the same hindlimb projectionfields in the sensorimotor cortex. For each of the 4 injec-tions, 0.5–1μl of 10% (w/v) BDA were injected using ananojet device (World Precision Instruments) at a depthof 1mm. The location of these injections is based on studiesshowing the location of CMN that project to the hindlimbsand forelimbs in Fischer rats [2] and confirmed by theexamination of the label in appropriate thalamic motornuclei. In some rats, gelfoam soaked in the BDA solutionwas placed over the exposed cortex prior to the closure ofthe craniotomy.

2.5. Functional Behaviour. A variety of behavioural tests forinjured rats were used to give a valid indication of functionalrecovery [24, 25, 32–35] including (i) “open-field” locomotion

test (BBB scoring method [36] to assess spontaneous move-ments), (ii) ladder walking [37] to assess general hindlimbrecovery, and (iii) our own novel computerized quantitativegait analysis method (Ratwalk® [38]) which allows objectiveanalyses of a large number of locomotion parameters, suchas interlimb coordination (step sequence), stride length, steplength, and base of support (stance width) [38]. All functional(locomotor) behaviour was assessed weekly for up to day 56after SCI. BBB scoring was analysed using at least 3 blindedraters both at the time of assessment and later in silica via slowmotion/frame-by-frame replay of high definition (1080 p)digital recordings taken during the test. The BBB rating scoreis composed of 22 nonlinear operational definitions (0–21scale) studying several aspects involved in the locomotion ofquadrupedal animals such as weight support, plantar step-ping, and forelimb/hindlimb coordination. Ladderwalk andRatwalk assays were performed on animals once they hadreached weight support on their hindlimbs, and involvedcomparison of preinjury performance with weekly assess-ments from day 14 post-SCI until day 56 (endpoint). Digitalrecordings of animals traversing 3 lengths per time point ofLadderwalk and Ratwalk (preinjury and weekly up to day 56post-SCI) were also similarly assessed by at least 2 blindedraters in silica via slow motion/frame-by-frame replay of dig-ital recordings taken during the tests for preinjury and up today 56 post-SCI. Ladderwalk scores are an average ofmisstepsover a 1m horizontal ladder with unevenly spaced bars.Ratwalk data analysis involves frame by frame designationfrom digital recordings of left and right fore- and hindlimbplacement on a 1m glass platform in the Ratwalk apparatusin low light, using an average of 3 “runs” per time point [38].

2.6. Perfusion and Tissue Processing. At day 56 (2 weeks afterthe BDA cortex injections), rats were euthanized by lethalinjection of sodium pentobarbitone (50mg/100 g) and trans-cardially perfused in 0.9M with 100ml of heparinized Dul-becco’s PBS followed by 4% (w/v) paraformaldehyde in PBSpH7.4. The head and vertebral column were dissected fromeach animal and postfixed for 24 hours. The brain and spinalcord were extracted from the skull and vertebra and thenstored intact in 0.1M PBS (pH7.4). The position of the injuryin the SC was measured from the caudal edge of the cerebel-lum to confirm that all animals were lesioned at the samelevel. A 2 cm segment was cut from the SC, with the lesionat the midpoint of this segment, and embedded in 1% (w/v)gelatin (Sigma-Aldrich). Using a CO2-freezing microtome(Polycut, Reichert-Jung, Australia), proximal and distal SCclose to the grafts (1 cm) was cut sagitally in 40μm frozensections, while the brain and brainstem were cut transverselyin 50μm frozen sections. A consecutive series of sectionswere transferred to 24-well plates containing 0.1M PB with0.01% (w/v) sodium azide (Sigma-Aldrich) and stored at4°C until processed for immunohistochemistry.

2.7. Tissue Analysis

2.7.1. Axonal Transduction and Anterograde Tracing. Frozencoronal brain sections were examined for preinjury AAV-CNTFmCherry or AAV-GFP transduction in the cortex and

3Neural Plasticity

(a) (b)

(d) (e)

(f) (g)

(i)

(h)

(j) (k) (l)

(c)

Figure 1: (a) Two AAV-CNTFmCherry injections in the cerebral cortex. (b) BDA injections into the cortex revealed using immunoperoxidase;note the many labelled neurons in deeper layers. This animal had previously received cortical injections of AAV-CNTFmCherry. (c–l)Longitudinal sections of the spinal cord—in all cases rostral is to the left of the picture. (c) Section immunostained for mCherry (red)and β-III tubulin (green) showing anterogradely labelled mCherry-positive axons (arrow) in the dorsal corticospinal tract far rostralto the lesion site. (d) Control AAV-GFP-injected rat (no MPCs injected); a small number of immunoperoxidase BDA-labelled axons anddebris are visible just in front of a rostral cyst (arrow), with no axons extending beyond the injury. (e–g) Large numbers of mCherry-positiveaxons rostral (e, f) and running dorsally over and beyond the cyst (g); these rats received AAV-CNTFmCherry cortical injections plus anintraspinal injection of viable rat mesenchymal precursor cells (rMPCs). Note in (e) the profusion of mCherry-positive profiles (large arrow)approximately 1mm rostral to the lesion cavity, growing into regions dorsal to the corticospinal tract (small arrows). There appears to beconsiderable sprouting of axons in this zone (f). (h) Immunoperoxidase BDA-labelled axons and debris rostral to a cyst in a rat injected withAAV-CNTFmCherry. Several axons can be seen running caudally, ventral to the cyst (arrows). (i) BDA-labelled cortical axons (arrows)visualized using a fluorescent secondary antibody (green), running over a cystic cavity. This animal also received AAV-CNTFmCherry andviable MPC injections. (j–l) Cortical axons (arrows) double labelled (l) with both mCherry ((j), red) and BDA ((k), green); note, some axonsare only mCherry or BDA immunoreactive. Scale bars: (a, e), 500μm; (b, h, and i), 200μm; (c, d, g, and j–l), 100μm; and (f) 50μm.

4 Neural Plasticity

thalamus using immunofluorescence, in addition to post-SCI anterograde BDA labelling using both immunofluores-cence and immunostaining with horseradish peroxidase(HRP). At the lesion site, longitudinal SC sections weresimilarly examined for axonal sprouting and regrowth(BDA). After blocking for 30min in PBS containing 10%(v/v) normal goat serum and 0.02% (v/v) Triton X-100,AAV-transduced axons were immunolabelled overnightat 4°C using antibodies to mCherry (Living Colours,1/600 in PBS). After washes, a secondary Cy3 goat anti-mouse antibody (Jackson ImmunoResearch 115613, 1/500dilution in PBS) was applied for 30min at room tempera-ture, before unbound antibody was washed away andsections coverslipped. In most cases, BDA was visualizedusing commercial VECTASTAIN avidin-biotin (VectorLaboratories, USA, number PK-4000) kits (as permanufacturer’s instructions) and horseradish peroxidase(HRP) histochemistry. Briefly, avidin-biotinylated (ABC)/HRP, followed by 3,3′-diaminobenzidine (DAB) solutionwas used to visualize BDA-stained axons. The ABC/HRPreagents were prepared 30 minutes before application to thesections. The sections were incubated with ABC reagents for1.5 hours at room temperature and washed with PBS. ADAB solution (10% (v/v) DAB metal concentrate in peroxi-dase buffer) was then added to the sections and incubated for5–15 minutes on a shaker. Sections were washed with PBS,allowed to air dry overnight and then counterstained with1% (w/v) toluidine blue and coverslipped with DEPEXmountingmedium (Fronine Lab Supply, Australia). To enableidentification of bothmCherry and BDA-labelled axons in thesame longitudinal tissue sections, BDAwas occasionally visu-alized using FITC-conjugated anti-streptavidin secondaryantibodies (Thermo Fisher Scientific). In AAV-GFP-injectedrats, brain and spinal cord sections were immunostainedwith an antibody to GFP (rabbit 1/500, Millipore, inPBS) followed by goat anti-rabbit IgG FITC-conjugatedsecondary antibodies (Jackson ImmunoResearch 89751,diluted 1/100 in PBS) similarly described as above.

In mCherry-, GFP-, and BDA-stained sections, theaccumulation of debris and branching of CST axons werecommonly seen rostral to the SCI and associated cysts.Axon sprouting was especially evident in AAV-CNTF-injected animals. In a series of sagittal sections, the rostralextent of this region was measured from the beginning of theinjury site in 38 rats (n = 6 for AAV-GFP; n = 15 for AAV-CNTFmCherry; n = 6 for AAV-CNTFmCherry +nonviablerMPCs; and n = 11 for AAV-CNTFmCherry + viable rMPCs).

2.7.2. Immunohistochemical Analyses of Glial and NeuronalPhenotypes. Brain and spinal cord tissue sections wereblocked in 10% (v/v) fetal calf serum (Gibco, BRL) and0.2% (v/v) Triton X-100 in PBS for 10 minutes at roomtemperature and washed in PBS. Primary antibodies (dilutedat 1/500 in PBS unless otherwise stated) to confirm theexpression of phenotypic markers for glial cells, GFAP inspinal cord (Millipore, AB3080), and axon populations usingantibodies to β-III tubulin (Covance, PRB-435) in the brainand spinal cord were used. Detection using FITC- or Cy3-conjugated secondary goat anti-mouse or goat anti-rabbit

antibodies (diluted at 1/400 in PBS, Jackson ImmunoRe-search, 115613, 89751) as described previously [24, 25].

2.7.3. Quantitative Analysis of Tissue Sparing in the SpinalCord. At least two independent raters (blinded) also mea-sured cyst sizes to remove bias. Tissue sparing was assessedby measuring cyst size and the amount of intact versusdegenerating tissue [24, 25]. Briefly, assessment of spinaltissue sparing was carried out using 0.05% (w/v) toluidineand 0.005% (w/v) borax solution followed by dehydrationin sequentially graded ethanol (v/v) of 70%, 90%, and100%. Staining on every sixth sagittal section was used todetermine the volume of spared spinal tissue. In each section,the total number of pixels in a 2.5mm-long SC segment wasdetermined, with the lesion epicenter in the middle, as well asthe area of damaged spinal tissue around it. The border of thedamaged tissue was defined by the absence of healthy cellsand an obvious discontinuity in density. Measurements ofeach section were summed per rat and averaged to give theamount of spared tissue, and percentage was calculated asthe difference between the area of damaged tissue versusthe whole segment (field of view) [24, 25].

2.7.4. Microscopy. A Nikon Eclipse E800 microscope wasused to visualise immunofluorescence staining as well as con-firm successful BDA injection into the brain and to visualizeBDA-stained axons at the SCI site. The distances between thefront edge of the most rostral cyst (if there was more thanone) and any axons observed alongside or caudal to the cystwere measured using the NIS elements BR 4.5 software.

2.7.5. Statistics. Using GraphPad Prism v4.03, SBSS (version21.0) and InStat v3.06 for Windows (GraphPad Software,San Diego, USA), 1- and 2-way repeated measures usingeither one- or two-way analysis of variance (ANOVA) plusTukey’s post hoc analysis as required were performed,except for BBB scoring which uses Kruskal-Wallis analysis(nonparametric ANOVA) as described previously [24, 25].In addition, Mann–Whitney post hoc testing was performed.

3. Results

3.1. Preinjury AAV Gene Therapy and Postinjury BDAInjection into the Cortex. Both GFP and mCherry expressionin transduced axons following AAV-GFP control and bicis-tronic AAV-CNTFmCherry injections into the cortex prior toSCI were confirmed at day 56 post-SCI using immunofluores-cence microscopy on relevant brain sections (Figure 1(a)).Note that the mCherry expression in neurons was confirmedthrough all the layers of cortex, including layer V containingCSN; this robust, post-2A linker expression is indicative ofwidespread transduction and expression of the secretableform of CNTF. BDA injection sites in the cortex werealso visualized with ABC/HRP and DAB reaction (seeFigure 1(b)) or immunofluorescence at day 56. BDAstaining reached layer V of the cortex indicating the success-ful labelling of CSN and, as for mCherry expression, the labelin the ventrolateral nucleus of the thalamus confirmedinjection into appropriate sensorimotor cortical regions.The appearance of BDA in the same tracts in the cortex as

5Neural Plasticity

transduced AAV-GFP control and AAV-CNTFmCherry

confirms that we were able to successfully label similar areasof axonal projection from the cortex that would facilitateidentification within the CST regions of the spinal cord.

3.2. AAV-Transduced and BDA-Labelled Projections in theSpinal Cord. In all treatment groups, AAV-transduced,immunostained axons and BDA-immunolabelled axons wereseen in the contralateral CST rostral to the injury site. Anexample is shown in Figure 1(c) with AAV-CNTFmCherry-labelled axons (red) in the CST adjacent to βIII tubulin-positivefibers andneurons (green).While nearly allCSTfiberswere well aligned in the ventral dorsal column in segments farrostral to the SCI site (Figure 1(c)), immediately in front of theinjury these axons became disorganized and more broadlydistributed. In this region, in addition to degenerate axon pro-files and other debris, apparently intact CST axons possessedcomplex, irregular profiles strongly suggestive of local sprout-ing and regenerative responses. In AAV-GFP control rats,there were small numbers of these axons immediately rostralto the first lesion cavity; however, in this group GFP-positiveor BDA-labelled CST axons were never seen caudal to theinjury/cyst. Furthermore, compared to animals injected withAAV-CNTFmCherry (Figures 1(e)–1(g)), there was greateraxon dieback as well as relatively little sprouting. The bestexample is shown in Figure 1(f) (BDA label). By comparison,in rats with AAV-CNTFmCherry cortical injections, therewas consistently a much higher density of CST axons400–1000μm rostral to the first spinal cord cavity, as revealedin bothmCherry- (Figures 1(e)–1(g)) and BDA- (Figure 1(h))immunostained material.

In a series of mCherry- or BDA-immunostained sagittalsections, the rostral extent of the zone containing scattered,often branched, CST axons was measured from the begin-ning of the injury site (taken as the rostral edge of the firstlesion cavity) in 38 rats. These “expanded” zones of CSTlabel (Figures 1(e)–1(h)) were seen in 3/6 AAV-GFP-injected rats, 9/15 AAV-CNTFmCherry rats, 4/6 rats withAAV-CNTFmCherry + nonviable rMPCs, and 8/11 rats withAAV-CNTFmCherry + viable rMPCs. The mean rostral extentof this zone was 323± 125μm (S.D.), 483± 226μm, 475± 330μm, and 638± 370μm for the four groups, respectively.There was considerable interanimal variability as shown bythe large standard deviations, nonetheless the trend forincreased density and greater areal extent of rostral CSTsprouting in AAV-CNTF-injected animals is evident.

After AAV-CNTF but not AAV-GFP cortical injections,many labelled CST axons were located beyond the rostraledge of the cyst (Figure 1(g)). Growth of BDA-positive axonsbeyond the cyst was also seen (arrows, Figure 1(h)). Long-distance growth of axons was also occasionally seen as shownin Figure 1(j); again, note the irregular nature of the postin-jury BDAFITC-labelled axonal profiles. This animal alsoreceived a viable MHC graft. In sections immunostained forboth mCherry (red) and BDA (FITC—green), we observedoccasional axons that were both AAV-CNTFmCherry- andanti-BDAFITC-positive (arrows, Figures 1(j)–1(l)). In theexample shown, the axons were located ventral to a cyst.The double labelling indicates that at last some of the

mCherry and BDA cortical injections were successfully madein overlapping regions of cortex, resulting in the dual label ofthe projecting layer V pyramidal neurons.

Figure 2 shows representative longitudinal spinalcord sections from three animals treated either withAAV-CNTFmCherry alone (Figures 2(A)–2(H)) or AAV-CNTFmCherry + viable rMPC transplantation (Figures 2(I),and 2(J)). Low power images show the size and location ofcysts in two of these animals (Figures 2(A) and 2(F), resp.).The arrows in Figures 2(A) and 2(F) point to the approxi-mate postcyst location of the mCherry-positive axonsshown in Figures 2(D), 2(E), and 2(I). In the rat shown inFigures 2(A)–2(E), there were large numbers of mCherry-positive fibers dorsal to the first large cyst in regions of sparedtissue midway within the lesion site (arrow, Figure 2(B),shown in higher power in Figure 2(C). More importantly,we also observed AAV-CNTFmCherry-positive fibers in the spi-nal cord distal to the lesion (Figures 2(D) and 2(E)). Thesefibers were nonlinear in orientation and, although infrequent,they were located up to 3-4mm beyond the rostral edge of theinjury site. In the animal shown in Figures 2(F)–2(H), againthere were many mCherry-immunopositive axons located justin front of, and dorsal to, the rostral cavity (Figure 2(G)), andmany sprouting BDA-positive axons were also seen. However,in this animal, there were only a small number of mCherry-labelled axons located caudal to the injury site (arrows,Figure 2(H)). It is worth emphasizing in Figures 2(D), 2(E),and 2(I) that the irregular course and branching of these axonsare strongly suggestive of regenerative growth. One other rat,this one treated with both AAV-CNTFmCherry and viablerMPCs, possessed a number of mCherry-labelled axons distalto the contusion injury and to the most caudal lesion cavity(arrows, Figures 2(I) and 2(J)).

3.3. Functional Hindlimb Recovery Is Generally Enhancedafter CNTF Gene Therapy

3.3.1. Ladderwalk. The number of missteps over theLadderwalk apparatus showed a gradual decline for alltreatment groups from day 14 through to endpoint atday 56 after SCI (Figure 3). Animals treated with controlAAV-GFP after SCI had the highest misstep counts acrossall time points, with the average decreasing gradually from12 at day14, to 8 at day 56, respectively. Rats with SCIfollowed by AAV-CNTFmCherry treatment consistentlydisplayed a lower average number of missteps than controlAAV-GFP treatment, decreasing gradually from ~9 atday14 to ~6 at day 56, respectively. The combined AAV-CNTFmCherry treatment with either nonviable or viablerMPC transplantation at the lesion site markedly reducedthe average numbers of missteps, ranging from 3 to 6 atday 14 after SCI, with both of these groups showing verysimilar average missteps of 3 from day 28 until day 56after SCI. Two-way ANOVA showed statistically signifi-cant differences (pairwise comparisons) between allgroups from day 14 until day 56 after SCI (p = 0 002,0.003, 0.006, 0.001, 0.013, 0.002, and 0.032 at days 14, 21, 28,35, 42, 49, and 56, resp.). Post hoc tests revealed statisticallysignificant differences between the AAV-GFP-treated control

6 Neural Plasticity

group and AAV-CNTFmCherry +nonviable rMPCs at all timepoints as well as between theAAV-GFP-treated control groupand AAV-CNTFmCherry + viable rMPCs at all time pointsexcept day 14 and day 56 (Figure 3).

Because the two-way ANOVA with repeated mea-sures described above showed a significant interaction(p = 0 0001), we then separated the factors and performed aone-way ANOVA on all AAV-CNTFmCherry-treated animals

Figure 2: Two examples (A–E and F–H) of animals that received cortical AAV-CNTFmCherry injections and with mCherry-positivecorticospinal axons distal to spinal cysts (D, E, I and J), and therefore distal to the initial injury. In all images rostral is to the left. (A, F)Low power views of cysts (β-III tubulin-immunostained sections) in each rat; the arrows in (A) and (B) point to the approximate locationof the axons shown in (D, E, and H), respectively. (B, C) Large numbers of mCherry-positive axons (arrow) dorsal to the large rostral cyst(see (A)), with small numbers of irregularly organized axons distal (D, E). (G) mCherry-labelled axons rostral and dorsal to the cyst, withseveral axons (arrows) located distal to the injury (H). In one animal injected with AAV-CNTFmCherry and that received a spinal rMPCinjection, numerous mCherry-positive axons (arrowed) were seen distal to the most caudal lesion cavity (I, J). Scale bars: (A, F), 1mm;(B), 500 μm; (C, G), 200 μm; (D, E, I, and J), 100μm; and (H) = 50μm.

7Neural Plasticity

at day 14 only. This one-way ANOVA revealed that there wasno difference between the cell-type (no cell, viable, and nonvi-able rMPC) treatments for the AAV-CNTFmCherry-treatedcohort at day 14 (p = 0 395), to be expected given that theLadderwalk tests were performed prior to rMPCtransplantation at this time point. Note that the number ofanimals per group at day 14was less than at later times becauseonly those animals that supported their weight on theirhindlimbs could be tested. Additionally, the one-way repeatedmeasures ANOVA within the AAV-CNTFmCherry animalsrevealed that while there was an overall effect of a reductionin missteps over time (p = 0 001), there was no differencebetween the cell-type treatment groups (p = 0 143). Interest-ingly however, in the AAV-CNTFmCherry-injected rats onlythe AAV-CNTFmCherry without rMPC group showed asignificant reduction in missteps (p = 0 05, LSD test) acrosstime with post hoc testing.

3.3.2. Open-Field Locomotion (BBB). BBB scores for hindlimbrecovery (Figure 4) revealed significant differences betweenthe AAV-GFP-treated control group (green) compared tothe AAV-CNTFmCherry (red, crosses), AAV-CNTFmCherry +nonviable rMPC (black), and AAV-CNTFmCherry + viablerMPC (blue) treatment groups, generally from day 21onwards. All treatment groups followed a slow increase inaverage BBB scores (0-1 at day 0 following SCI) until aroundday 5–day 7 when average scores began to typically increaseat a greater rate, with the AAV-GFP-treated control group

consistently showing the lowest average BBB scores of 5 fromday 7 (which corresponds to a slight movement of two jointsand an extensive movement of the third joint) until score10 at day 56 after SCI (which corresponds to plantarstepping with occasional weight bearing and no forelimb-hindlimb coordination). AAV-CNTFmCherry and AAV-CNTFmCherry + nonviable rMPC groups’ scores plateauedfrom day 21 after SCI with an average BBB score of 11 (whichcorresponds to plantar stepping with frequent to consistentweight bearing and NO forelimb-hindlimb coordination).Only the AAV-CNTFmCherry and AAV-CNTFmCherry + viablerMPC treatment group obtained higher scores from day 42until day 56 after SCI, where average BBB scores plateauedat 12 (which corresponds to plantar stepping with frequentto consistent weight bearing and occasional forelimb-hindlimb coordination), suggesting that CNTF gene therapyalone and CNTF gene therapy+ viable MPC transplantationinto the lesion resulted in the best functional outcomes.Kruskal Wallis scores showed overall statistically significantdifferences between the AAV-GFP-treated control groupand all other treatment groups at day 21 (∗∗p = 0 006),day 28 (∗∗∗p = 0 0016), and day 49 (∗p = 0 023). SpecificMann–Whitney post hoc tests revealed statistically signifi-cant differences between the AAV-GFP-treated control groupand AAV-CNTFmCherry (p = 0 001 at day 21, p = 0 004 at day28, and p = 0 003 at day 49), AAV-CNTFmCherry +nonviablerMPCs (p = 0 018 at day 21, p = 0 036 at day 28, andp = 0 113 at day 49), and AAV-CNTFmCherry + viable

18

16

14

12

10

8

6

4Aver

age n

umbe

r of m

isste

ps

2

014 21 28 35

Days after SCI42 49 56

AAV-GFP controlAAV-CNTF only

AAV-CNTF + n.v. cellsAAV-CNTF + viable cells

⁎⁎⁎ ⁎⁎⁎ ⁎⁎ ⁎⁎⁎ ⁎ ⁎⁎⁎

Figure 3: Functional hindlimb recovery promoted after AAV-CNTF gene therapy and cellular transplantation as assessed by Ladderwalk:reduced misstepping over the Ladderwalk apparatus indicates that AAV-CNTFmCherry therapy (red, crosses) promoted significantfunctional improvement after SCI compared with control AAV-GFP treatment (green). Note that at day 14, the Ladderwalk testing wascarried out prior to injection of either viable (blue) or nonviable (black) cells. Post hoc tests revealed statistically significant differencesbetween the AAV-GFP-treated control group and AAV-CNTFmCherry + nonviable rMPCs at all time points as well as between the AAV-GFP-treated control group and AAV-CNTFmCherry + viable rMPCs at all time points except day 56 (∗p = 0 01 – 0 05, ∗∗p = 0 005 – 0 01,and ∗∗∗p = 0 001 – 0 005). Two-way repeated measures ANOVA was conducted using PRISM (time: p = 0 0001, treatment: p = 0 0016,and interaction: p = 0 0001). Standard deviation is shown.

8 Neural Plasticity

rMPCs (p = 0 021 at day 21, p = 0 008 at day 28, andp = 0 059 at day 49), respectively.

3.3.3. Ratwalk® Gait Analysis. Ratwalk analysis on animalsshowing hindlimb weight support are summarized inFigure 5, with averages of arbitrary unit values assigned bythe software in silica for each treatment group compared topreinjury levels (dotted black lines) shown for stance width(5A), step length (5B), stride length (5C), and step sequence(5D) at day 56 after injury, by which time any differencesin gait parameters should be apparent. Stance width is theaverage distance between each of the forelimbs (Rf/Lf), eachof the hindlimbs (Rh/Lh), and each of the fore-to-hindlimbplacements (Rf/Lh and Lf/Rh). There was no significantdifference between any treatment group for any variable instance width at day 56. Typically, hindlimb stance widthincreases slightly very early on after SCI (data not shown)and although the AAV-GFP-treated control group stillshowed a marginally increased average hindlimb stancewidth compared to all other treatment groups, there was nostatistical difference between them or preinjury levels. Ahigher average stance width between opposing fore- andhindlimbs (Rf/Lh and Lf/Rh) was maintained at day 56post-SCI for all treatment groups compared to preinjurylevels and suggests that all animals assume a longer distancebetween each fore- and hindlimbs as a consequence of theinjury irrespective of treatment. The stance width datasuggesting these compensatory fore- and hindlimb changesin the treatment groups is confirmed by the day 56 post-SCI analyses for step length (Figure 5(b)) and stride length(Figure 5(c)), which again showed no statistically significant

differences between treatments. Step length analysis showsa higher average distance between Rf/Rh, Rf/Lh, Rh/Lf, andLf/Lh across all groups compared to preinjury levels withno statistically significant differences between the groups(Figure 5(b)). Stride length also shows that generally forelimbstrides are shorter and hindlimb strides are longer on averagecompared to preinjury levels (Figure 5(c)), consistent withthe idea that compensatory movements to bring forelimbscloser and take shorter strides to help “pull” the animalforward correlates with longer hindlimb strides that adopt awider stance (for more stability) and therefore longerdistances. While the AAV-CNTFmCherry + nonviable rMPCgroup of animals still showed slightly higher forelimb dis-tances compared to preinjury levels, there were no statisti-cally significant differences between treatments.

Step sequence analysis revealed that following SCI, therewas generally a decrease in the amount of patterns of coordi-nated fore- and hindlimb placements designated as “cruci-ate,” “alternate,” and “rotary” using the Ratwalk software,as well as an increase in the amount of unrecognisable (non-coordinated) fore- and hindlimb placements designated as“none” (Figure 5(d)). There remained a significant decreasein the amount of cruciate patterns of sequence comparedto preinjury levels, and despite some variation betweentreatment groups, there were no statistically significantdifferences between treatment groups for both alternateand rotary patterns of step sequence. The amount ofuncoordinated patterns of step sequence (“none”) wasalmost twice as high for all treatment groups at day 56post-SCI, again with no statistically significant differencesbetween treatment groups.

14

AAV-GFP controlAAV-CNTF only

AAV-CNTF + nonviable cellsAAV-CNTF + viable cells

12

10

8

6BBB

scor

e

4

2

01 3 5 7 14 21

Day after SCI28 35 42 49 56

⁎⁎⁎ ⁎ ⁎

Figure 4: Functional hindlimb recovery promoted after AAV-CNTF gene therapy and cellular transplantation as assessed by open-fieldlocomotion (BBB). Significant functional improvements were observed following AAV-CNTFmCherry (red), AAV-CNTFmCherry + nonviablerMPC (black), and AAV-CNTFmCherry + viable rMPC (blue) treatment compared to AAV-GFP-treated control animals (green) generallyfrom day 21 after SCI onwards, although statistically significant differences were not maintained at all time points (∗p = 0 01 – 0 05and ∗∗∗p = 0 001 – 0 005). Standard deviation is shown.

9Neural Plasticity

3.4. Transplantation of Viable rMPCs, but Not Cortical GeneTherapy, Promotes Tissue Sparing. Cysts developed at theinjury site of all animals in all groups after SCI. The cyst sizesmeasured at day 56 after SCI are shown in Figure 6. AAV-GFP treatment and AAV-CNTFmCherry treatment groupson average had a 10-11% area of field of view occupied bycysts, respectively. While AAV-CNTFmCherry + nonviablerMPC-treated animals still showed slightly lower average cystsizes of about 8%, there was no statistical difference betweenany of these groups. However, AAV-CNTFmCherry transduc-tion in the cortex followed by viable rMPC transplantationinto the lesion did result in a statistically significant reductionin average cyst size, to just a 2% area of field of view, indicat-ing that viable rMPCs are able to significantly alter the terrainof the lesion site to affect tissue sparing. Note that two of thethree rats with mCherry-positive axons distal to the injurysite were in the AAV-CNTFmCherry + viable rMPC group.Donor rMPCs were not labelled with any marker foridentification posttransplantation, and there was no evidenceof individual donor rMPCs remaining in spinal cord

sections subjected to immunohistochemistry or toluidineblue staining (data not shown).

4. Discussion

In this study, we used AAV-CNTFmCherry therapy to trans-duce neurons, including CSN, in the sensorimotor cortex ofanimals with a moderate thoracic (T10) contusion injury,with the aim of enhancing plasticity and promoting theregrowth of corticospinal tract axons after SCI. It is widelyacknowledged that combinations of therapies are requiredfor effective treatment of SCI, thus in some animals AAV-CNTFmCherry was applied in combination with viable ornonviable rat mesenchymal precursor cells (rMPCs) graftedinto the spinal lesion site two weeks after SCI to modulatethe local inhibitory environment. As discussed more fullybelow, different types of AAV vectors have previously beeninjected into the cortex in SCI studies, and MPC grafts havealso been tested after SCI, but to our knowledge no trialshave—until now—combined these therapeutic approaches.

120.00

100.00

80.00

60.00

40.00

20.00

0.00RF/Lf Rh/Lh

Stance width day 56

RF/Lf Rh/Lh

AAV-GFP controlAAV-CNTF only

AAV-CNTF + n.v. cellsAAV-CNTF + viable cells

(a)

Step length day 56120.00

100.00

80.00

60.00

40.00

20.00

0.00Rf/Rh Rf/Lh Rh/Lf Lf/Lh

AAV-GFP controlAAV-CNTF only

AAV-CNTF + n.v. cellsAAV-CNTF + viable cells

(b)

Stride length day 56120.00

100.00

80.00

60.00

40.00

20.00

0.00Rf Rh Lf Lh

AAV-GFP controlAAV-CNTF only

AAV-CNTF + n.v. cellsAAV-CNTF + viable cells

(c)

Step sequence day 56120.00

100.00

80.00

60.00

40.00

20.00

0.00RF/Lf Rh/Lh RF/Lf Rh/Lh

LF

LH RH

RF

AAV-GFP controlAAV-CNTF only

AAV-CNTF + n.v. cellsAAV-CNTF + viable cells

(d)

Figure 5: Ratwalk gait analysis at day 56 after SCI. Averages of arbitrary unit values of each treatment group compared to preinjury levels(dotted black lines) are shown for stance width (a), step length (b), stride length (c), and step sequence (d). Despite no statisticallysignificant differences between any treatment group for any variable, compensatory changes such as a marked reduction in the stancewidth (a) of the forelimbs (Rf/Lf) between opposing fore- and hindlimbs (Rf/Lh and Lf/Rh) in all groups compared to preinjury levelswere supported by data for step length (b) and stride length (c). A general decrease in the amount of patterns of coordinated fore- andhindlimb placement (“cruciate,” “alternate,” and “rotary” as indicated in (d)) and an increase in the amount of noncoordinated fore- andhindlimb placement (“none”) was observed. Standard deviation is shown.

10 Neural Plasticity

AAV-CNTFmCherry therapy alone or in combination witheither viable or nonviable rMPC transplantation provided asustained improvement in functional outcome over AAV-GFP alone as measured by Ladderwalk. Note that with thisbehavioural test, while control and CNTF treatmentsdiffered, there was no significant difference between thedifferent AAV-CNTFmCherry groups (no cells, viable cells,and nonviable cells), thus it seems that the presence ofcortical CNTF was sufficient to yield an improvement inthe stepping function. AAV-CNTFmCherry therapy aloneand in combination with rMPC transplantation also yieldedsustained improvements as assessed by open-field locomo-tion (BBB). Ratwalk gait analyses revealed subtle compen-satory mechanisms for limb placement after injury, but itis likely that even the moderate injury used here was toosevere to reveal significant statistical differences. The aver-age cyst size was significantly reduced only in the AAV-CNTFmCherry + viable rMPC group.

In addition to many degenerative profiles, the sproutingof apparently intact axons was evident in the CST immedi-ately rostral to the injury site. Compared to AAV-GFP con-trols, considerably more BDA or mCherry-labelled axonswith complex profiles were located just rostral to the cystsin the AAV-CNTFmCherry-injected rats containing CNTF-transduced cortical neurons. Not only was the density ofthese axons higher, but the area of the spinal cord containingthem extended further rostrally into uninjured segments.Furthermore, only in AAV-CNTFmCherry-transduced treat-ment groups were BDA and mCherry-immunoreactiveaxons seen to project alongside, and sometimes severalmillimeters beyond, the rostral border of the lesion-inducedcysts. It is important to emphasize that, compared to orientedand organized fibers in the CST far rostral to the injury site,axons were more irregularly organized immediately in frontof, as well as caudal to, cysts, strongly suggestive of localsprouting and regenerative responses.

In three of the eleven animals that received cortical AAV-CNTFmCherry injections, axons were seen caudal to all cysts,located up to 3-4mm beyond the rostral edge of the injurysite. These axons were seen in two non-rMPC grafted ani-mals as well as in a rat that also received an rMPC transplant.We did not detect such axons in AAV-GFP-injected animals,again supporting the suggestion that CNTF has the potentialto enhance axon regeneration in the spinal cord by transduc-ing appropriate neuronal populations in the sensorimotorcortex. In other CNS systems, vector-mediated delivery ofCNTF to parent cell bodies combined with the transplanta-tion of an appropriate bridging substrate (peripheral nerve),promotes the long-distance regeneration of injured axons[7–10], with at least some functional outcomes [11]. Notein particular that the survival effects of CNTF have beendemonstrated on numerous neuron populations [14–16],including increasing the viability of corticospinal neuronsin the sensorimotor cortex after intracortical axotomy[6]. Because AAV-CNTF delivery is only to the cortex,the primary site of trophic action is at the cell bodies,and transport and release of CNTF at axon tips likelyoccurs at a low level (we have confirmed this previously,data not shown). Based on previous studies in the visualsystem, we propose that the mechanisms of protectionand promotion of plasticity and regenerative capacity aremediated and amplified both by direct infection of CSNthemselves but also due to bystander release of CNTF byadjacent nonprojecting cortical cells [8, 39]. Many of thebiological actions of CNTF are signalled via STAT3, andit is therefore important to note that vector-mediateddelivery of a hyperactivated STAT3 enhances processoutgrowth in cultured cortical neurons [40], and in vivooverexpression of STAT3 enhances CSN plasticity in miceafter SCI [41].

Delivery of AAV vectors to the cortex has been reportedby a number of groups, some to test the optimal serotype for

AAV-GFP control AAV-CNTF only AAV-CNTF +nonviable cells

AAV-CNTF + viable cells0.00

2.00

4.00

Aver

age %

cyst

area

(F.O

.V.)

6.00

8.00

10.00

12.00

⁎

Figure 6: Tissue sparing promoted by combined cortical AAV-CNTFmCherry therapy and local transplantation of viable rMPCs. AAV-GFPcontrol treatment and AAV-CNTFmCherry treatment groups revealed similar average areas of cyst formation in spinal cord sections stained bytoluidine blue, and although AAV-CNTFmCherry + nonviable rMPC treatment resulted in slightly lower average cyst sizes, onlyAAV-CNTFmCherry + viable rMPC treatment into the lesion resulted in a statistically significant reduction in average cyst size. ∗p = 0 01 – 0 05.

11Neural Plasticity

transduction of cortical neurons, including CSN [42, 43],others to deliver factors designed to enhance CST repair afterSCI [44–49]. Such studies are nonetheless much less frequentthan those involving the delivery of vectors and/or growthfactors to the spinal cord injury site itself [2, 50]. In someinitial studies, we tested the effect of the cortical injectionof AAV1 expressing insulin-like growth factor 1 (IGF-1) onCST plasticity after SCI, but although we saw some additionalsprouting rostral to the injury, this was less than that seenafter the cortical delivery of AAV-CNTF vectors, and wesaw no significant changes in functional recovery (unpub-lished data). Interestingly then, postlesion AAV-assistedcoexpression of IGF1 and osteopontin in cortical neuronsresulted in robust CST regrowth and the recovery ofCST-dependent behavioural outcomes after SCI [50], againindicative of the therapeutic power of combined therapies.

The long-term transduction of murine cortical neuronsusing an AAV vector to suppress or conditionally deleteexpression of phosphatase and tensin homolog (PTEN), themain suppressor of the PI3K-Akt survival and growthpathway, leads to the greater regenerative growth of CSTaxons and some functional improvements [45]. Howeverthe resultant long-term upregulation of protein kinases suchas mTOR, a key regulator of protein translation in neurons,leads to some aberrant growth [48] and progressive changesin the growth of cells, their dendrites, and their axons [46].This is not dissimilar to the effect of long-term (up toone year) AAV-mediated expression of CNTF on retinalganglion cells, where there are not only changes in thedendritic morphology of transduced and neighboring non-transduced cells [51], but there is also altered expressionof endogenous retinal genes [52]. Whether such effects arealso seen in cortical neurons transduced with AAV-CNTFis yet to be determined, but this is an important topic forfuture studies.

MPCs from the bone marrow stroma have the potentialto differentiate into cells with many of the phenotypiccharacteristics of neural tissue [53–55], migrate and inte-grate into CNS tissues, and express markers typical ofmature neurons and astrocytes [56]. MPCs have beensuccessfully transplanted into the spinal cord and shownto (a) promote regeneration of lesioned axons into thegraft, (b) differentiate into neurons, (c) remyelinate dam-aged myelin sheaths around CNS axons, and (d) improvefunctional outcomes after SCI (for extensive reviews, see[57, 58]). Our own work using purified (Stro-1+) humanMPCs from the bone marrow stroma of SCI patientsdramatically improved anatomical (characterized by smallercyst sizes, as well as lower amounts of degenerative tissue)and functional recovery after both acute and subacute/chronic SCI in nude rat (T-cell immunodeficient) hosts aftercontusion SCI [24, 25]. Rat MPCs can be prepared to anessentially pure, minor subpopulation of adult cells [59–61]significantly lessening any potential variation in functionaland morphological outcomes, often encountered in theliterature when using different hMPC donor cells in trans-plantation experiments.

rMPC transplantation at day 14 post-SCI presumablyavoids the cytotoxicity of the acute injury and presumably

allows sufficient time for targeted expression of CNTF to beswitched on in transduced cortical cells [2, 4, 28]. Generally,SCI studies show significant results attributable to transplan-tation within a few weeks of injection [24, 25, 57, 62, 63];however, in the present study the impact of rMPCs graftedinto the injury site was evident in reducing postlesion cavitydimensions, but effects were absent or inconsistent in thebehavioural studies involving AAV-CNTFmCherry-injectedanimals. It should be noted that a caveat of this study is thatbehavioural tests were continued after BDA was injectedwhich could potentially obscure or increase treatment effectsdue to the injection-associated injury and surgery, althoughwe did not observe any apparent reduction in functionaloutcomes between BDA injection and the final time pointsanalysed. There did not appear to be greater macrophage/microglial reactivity associated with AAV transduction(either AAV-GFP-treated control or AAV-CNTFmCherry

treatments) beyond that typically observed in our other SCIstudies [24, 25]. MPCs do produce some CNTF [64] butour experience is that, while MPCs alter the local environ-ment to enhance axonal regrowth, few survive in the longterm [24, 25]. Only viable rMPC grafts promoted statisticallysignificant tissue sparing, yet nonviable rMPC transplanta-tion had similar effects as viable rMPC transplantation infunctional outcomes (as shown in our Ladderwalk andopen-field assays). The use of (freeze-thawed) nonviable cellsas appropriate controls for cell transplantation studies in SCIis by no means common. There is evidence to suggest thateven fibroblasts can contribute to functional and/or morpho-logical improvements in animal models of SCI compared tospecific stem/precursor cell types [57, 63]. Indeed somestudies have reported functional improvements withoutassociated (and expected?) structural improvements, andvice versa [57, 65]. A possible scenario is that MPCs (whichare known to have immunosuppressive properties [66])may act as immune “decoys” that modulate the host immuneresponse (e.g., [67]) and this property may still be effectiveeven if the cells themselves are not viable. This couldpotentially allow the “normal” host repair mechanism to bemore effective and be reflected in either functional and/ormorphological outcomes.

4.1. Impact and Future Direction. The combination of AAV-targeted expression of CNTF, with or without the use of stemcell graft technology, represents a novel strategy to assess theeffect of vector-mediated production of growth factors onplasticity and regeneration after SCI. A major aim of theseexperiments was to assess the capability of cortical genetherapy to promote potential plasticity and regrowth/regeneration of CST axons. The present behavioural andmorphological data after thoracic contusion injury showthe promise of using cortical AAV-CNTF gene therapyto promote repair after SCI. In future studies, we will testthis approach in a model of cervical CST hemitransection,focusing on the importance of this tract in rodent forelimbfunction. The data presented here aid in the advancement oftechnologies related to the development of more effectivegene therapy and begin to provide a platform for exploringthe possibility of preclinical studies aimed at using gene

12 Neural Plasticity

therapy to modify cortical neurons as part of an SCIrepair strategy.

Data Availability

The histological slides are stored in Stuart I. Hodgetts’laboratory, as are videos of behavioural analyses.

Conflicts of Interest

The authors declare that they have no conflicts of interest.

Acknowledgments

The authors thank Associate Professor Peter Mark forstatistical help and advice,Mary Lee for laboratory assistanceand tissue processing, and Guy Ben-Ary for image analysis.Additionally, the authors thank Professor Giles Plant and IainSweetman (Ratwalk) for the development of the Ratwalkdevice and software. Professor Giles Plant is affiliated withthe Stanford Partnership for Spinal Cord Injury andRepair, Stanford Institute for Neuro-Innovation and Transla-tional Neurosciences, and Department of Neurosurgery,265 Campus Drive, Stanford, CA 94305-5454, USA. IainSweetman (Ratwalk) is affiliated with the Faculty, Preventiveand Social Medicine NZPhvC Health Sciences, DunedinSchool of Medicine, Otago, New Zealand.

References

[1] C. Watson and A. R. Harvey, “Chapter 11. Projectionsfrom the brain to the spinal cord,” in The Spinal Cord,pp. 168–179, Academic Press, 2009.

[2] A. R. Harvey, S. J. Lovett, B. T. Majda, J. H. Yoon, L. P. G.Wheeler, and S. I. Hodgetts, “Neurotrophic factors for spinalcord repair: which, where, how and when to apply, and forwhat period of time?,” Brain Research, vol. 1619, pp. 36–71,2015.

[3] E. J. Gonzalez-Rothi, A. M. Rombola, C. A. Rousseau et al.,“Spinal interneurons and forelimb plasticity after incompletecervical spinal cord injury in adult rats,” Journal of Neuro-trauma, vol. 32, no. 12, pp. 893–907, 2015.

[4] T. Isa and Y. Nishimura, “Plasticity for recovery after partialspinal cord injury—hierarchical organization,” NeuroscienceResearch, vol. 78, pp. 3–8, 2014.

[5] C. Darian-Smith, A. Lilak, J. Garner, and K. A. Irvine,“Corticospinal sprouting differs according to spinal injurylocation and cortical origin in macaque monkeys,” The Journalof Neuroscience, vol. 34, no. 37, pp. 12267–12279, 2014.

[6] S. M. Dale, R. Z. Kuang, X. Wei, and S. Varon, “Corticospinalmotor neurons in the adult rat: degeneration after intracorticalaxotomy and protection by ciliary neurotrophic factor(CNTF),” Experimental Neurology, vol. 135, no. 1, pp. 67–73,1995.

[7] A. R. Harvey, Y. Hu, S. G. Leaver et al., “Gene therapy andtransplantation in CNS repair: the visual system,” Progressin Retina and Eye Research, vol. 25, no. 5, pp. 449–489,2006.

[8] M. Hellstrom, M. A. Pollett, and A. R. Harvey, “Post-injurydelivery of rAAV2-CNTF combined with short-term pharma-cotherapy is neuroprotective and promotes extensive axonal

regeneration after optic nerve trauma,” Journal of Neuro-trauma, vol. 28, no. 12, pp. 2475–2483, 2011.

[9] Y. Hu, S. G. Leaver, G. W. Plant et al., “Lentiviral-mediatedtransfer of CNTF to Schwann cells within reconstructedperipheral nerve grafts enhances adult retinal ganglion cellsurvival and axonal regeneration,” Molecular Therapy,vol. 11, no. 6, pp. 906–915, 2005.

[10] S. G. Leaver, Q. Cui, O. Bernard, and A. R. Harvey, “Coopera-tive effects of bcl-2 and AAV-mediated expression of CNTF onretinal ganglion cell survival and axonal regeneration in adulttransgenic mice,” The European Journal of Neuroscience,vol. 24, no. 12, pp. 3323–3332, 2006.

[11] S. W. You, M. Hellström, M. A. Pollett et al., “Large-scalereconstitution of a retina-to-brain pathway in adult ratsusing gene therapy and bridging grafts: an anatomical andbehavioral analysis,” Experimental Neurology, vol. 279,pp. 197–211, 2016.

[12] N. Y. Ip and G. D. Yancopoulos, “Ciliary neurotrophic factorand its receptor complex,” Progress in Growth Factor Research,vol. 4, no. 2, pp. 139–155, 1992.

[13] N. Y. Ip and G. D. Yancopoulos, “The neurotrophins andCNTF: two families of collaborative neurotrophic factors,”Annual Review of Neuroscience, vol. 19, no. 1, pp. 491–515,1996.

[14] P. M. Richardson, “Ciliary neurotrophic factor: a review,”Pharmacology & Therapeutics, vol. 63, no. 2, pp. 187–198,1994.

[15] M. Sendtner, P. Carroll, B. Holtmann, R. A. Hughes, andH. Thoenen, “Ciliary neurotrophic factor,” Journal of Neurobi-ology, vol. 25, no. 11, pp. 1436–1453, 1994.

[16] M. Sendtner, F. Dittrich, R. A. Hughes, and H. Thoenen,“Actions of CNTF and neurotrophins on degeneratingmotoneurons: preclinical studies and clinical implications,”Journal of the Neurological Science, vol. 124, pp. 77–83,1994.

[17] J. Ye, L. Cao, R. Cui et al., “The effects of ciliary neurotrophicfactor on neurological function and glial activity followingcontusive spinal cord injury in the rats,” Brain Research,vol. 997, no. 1, pp. 30–39, 2004.

[18] R. E. Clatterbuck, D. L. Price, and V. E. Koliatsos, “Ciliaryneurotrophic factor prevents retrograde neuronal death inthe adult central nervous system,” Proceedings of the NationalAcademy of Sciences of the United States of America, vol. 90,no. 6, pp. 2222–2226, 1993.

[19] T. Hagg and S. Varon, “Ciliary neurotrophic factor preventsdegeneration of adult rat substantia nigra dopaminergicneurons in vivo,” Proceedings of the National Academy ofSciences of the United States of America, vol. 90, no. 13,pp. 6315–6319, 1993.

[20] P. H. Ozdinler and J. D. Macklis, “IGF-I specifically enhancesaxon outgrowth of corticospinal motor neurons,” NatureNeuroscience, vol. 9, no. 11, pp. 1371–1381, 2006.

[21] Q. Cao, Q. He, Y. Wang et al., “Transplantation of ciliaryneurotrophic factor-expressing adult oligodendrocyte precur-sor cells promotes remyelination and functional recovery afterspinal cord injury,” The Journal of Neuroscience, vol. 30, no. 8,pp. 2989–3001, 2010.

[22] A. R. Harvey, J. W. Ooi, and J. Rodger, “Chapter one -Neurotrophic factors and the regeneration of adult retinalganglion cell axons,” International Review of Neurobiology,vol. 106, pp. 1–33, 2012.

13Neural Plasticity

[23] J. L. Nielson, I. Sears-Kraxberger, M. K. Strong, J. K. Wong,R. Willenberg, and O. Steward, “Unexpected survival ofneurons of origin of the pyramidal tract after spinal cordinjury,” The Journal of Neuroscience, vol. 30, no. 34,pp. 11516–11528, 2010.

[24] S. Hodgetts, P. Simmons, and G. W. Plant, “Human mesen-chymal precursor cells (Stro-1+) from spinal cord injurypatients improve functional recovery and tissue sparing in anacute spinal cord injury rat model,” Cell Transplantation,vol. 22, no. 3, pp. 393–412, 2013.

[25] S. Hodgetts, P. Simmons, and G. W. Plant, “A comparison ofthe behavioral and anatomical outcomes in sub-acute andchronic spinal cord injury models following treatment withhuman mesenchymal precursor cell transplantation andrecombinant decorin,” Experimental Neurology, vol. 248,pp. 343–359, 2013.

[26] E. D. Wirth 3rd, P. J. Reier, R. G. Fessler et al., “Feasibility andsafety of neural tissue transplantation in patients with syringo-myelia,” Journal of Neurotrauma, vol. 18, no. 9, pp. 911–929,2001.

[27] W. Tang, I. Ehrlich, S. B. E. Wolff et al., “Faithful expression ofmultiple proteins via 2A-peptide self-processing: a versatileand reliable method for manipulating brain circuits,” TheJournal of Neuroscience, vol. 29, no. 27, pp. 8621–8629, 2009.

[28] H. Petrs-Silva and R. Linden, “Advances in recombinantadeno-associated viral vectors for gene delivery,” Current GeneTherapy, vol. 13, no. 5, pp. 335–345, 2013.

[29] T. H. Hutson, J. Verhaagen, R. J. Yanez-Munoz, and L. D.Moon, “Corticospinal tract transduction: a comparison ofseven adeno-associated viral vector serotypes and a non-integrating lentiviral vector,” Gene Therapy, vol. 19, no. 1,pp. 49–60, 2012.

[30] M. Sendtner, H. Schmalbruch, K. A. Stöckli, P. Carroll, G. W.Kreutzberg, and H. Thoenen, “Ciliary neurotrophic factorprevents degeneration of motor neurons in mouse mutantprogressive motor neuronopathy,” Nature, vol. 358, no. 6386,pp. 502–504, 1992.

[31] A. Harvey, S. Clarkson, and G. Plant, “Corticospinal projec-tions in adult Fischer F344 rats,” Proceedings of the AustralianNeuroscience Society, vol. 16, p. 103, 2005.

[32] F. P. Hamers, G. C. Koopmans, and E. A. Joosten,“CatWalk-assisted gait analysis in the assessment of spinalcord injury,” Journal of Neurotrauma, vol. 23, no. 3-4,pp. 537–548, 2006.

[33] F. P. Hamers, A. J. Lankhorst, T. J. van Laar, W. B. Veldhuis,and W. H. Gispen, “Automated quantitative gait analysisduring overground locomotion in the rat: its application tospinal cord contusion and transection injuries,” Journal ofNeurotrauma, vol. 18, no. 2, pp. 187–201, 2001.

[34] G. D. Muir and A. A. Webb, “Assessment of behaviouralrecovery following spinal cord injury in rats,” EuropeanJournal of Neuroscience, vol. 12, no. 9, pp. 3079–3086, 2000.

[35] G. W. Plant, P. F. Currier, E. P. Cuervo et al., “Purified adultensheathing glia fail to myelinate axons under cultureconditions that enable Schwann cells to form myelin,” TheJournal of Neuroscience, vol. 22, no. 14, pp. 6083–6091, 2002.

[36] D. M. Basso, M. S. Beattie, and J. C. Bresnahan, “A sensitiveand reliable locomotor rating scale for open field testing inrats,” Journal of Neurotrauma, vol. 12, no. 1, pp. 1–21, 1995.

[37] G. A. Metz and I. Q. Whishaw, “Cortical and subcorticallesions impair skilled walking in the ladder rung walking test:

a new task to evaluate fore- and hindlimb stepping, placing,and co-ordination,” Journal of Neuroscience Methods,vol. 115, no. 2, pp. 169–179, 2002.

[38] M. J. Godinho, L. Teh, M. A. Pollett et al., “Immunohisto-chemical, ultrastructural and functional analysis of axonalregeneration through peripheral nerve grafts containingSchwann cells expressing BDNF, CNTF or NT3,” PLoS One,vol. 8, no. 8, article e69987, 2013.

[39] S. G. Leaver, Q. Cui, G. W. Plant et al., “AAV-mediatedexpression of CNTF promotes long-term survival and regener-ation of adult rat retinal ganglion cells,” Gene Therapy, vol. 13,no. 18, pp. 1328–1341, 2006.

[40] S. T. Mehta, X. Luo, K. K. Park, J. L. Bixby, and V. P. Lemmon,“Hyperactivated Stat3 boosts axon regeneration in the CNS,”Experimental Neurology, vol. 280, pp. 115–120, 2016.

[41] C. Lang, P. M. Bradley, A. Jacobi, M. Kerschensteiner, andF. M. Bareyre, “STAT3 promotes corticospinal remodellingand functional recovery after spinal cord injury,” EMBOReports, vol. 14, no. 10, pp. 931–937, 2013.

[42] D. F. Aschauer, S. Kreuz, and S. Rumpel, “Analysis oftransduction efficiency, tropism and axonal transport ofAAV serotypes 1, 2, 5, 6, 8 and 9 in the mouse brain,” PLoSOne, vol. 8, no. 9, article e76310, 2013.

[43] A. Watakabe, M. Ohtsuka, M. Kinoshita et al., “Comparativeanalyses of adeno-associated viral vector serotypes 1, 2, 5, 8and 9 in marmoset, mouse and macaque cerebral cortex,”Neu-roscience Research, vol. 93, pp. 144–157, 2014.

[44] N. Weishaupt, S. Li, A. Di Pardo, S. Sipione, and K. Fouad,“Synergistic effects of BDNF and rehabilitative training onrecovery after cervical spinal cord injury,” Behavioural BrainResearch, vol. 239, pp. 31–42, 2013.

[45] C. A. Danilov and O. Steward, “Conditional genetic deletion ofPTEN after a spinal cord injury enhances regenerative growthof CST axons and motor function recovery in mice,” Experi-mental Neurology, vol. 266, pp. 147–160, 2015.

[46] E. A. Gallent and O. Steward, “Neuronal PTEN deletion inadult cortical neurons triggers progressive growth of cellbodies, dendrites, and axons,” Experimental Neurology,vol. 303, pp. 12–28, 2018.

[47] E. A. Gutilla, M. M. Buyukozturk, and O. Steward, “Long-termconsequences of conditional genetic deletion of PTEN in thesensorimotor cortex of neonatal mice,” Experimental Neurol-ogy, vol. 279, pp. 27–39, 2016.

[48] R. Willenberg, K. Zukor, K. Liu, Z. He, and O. Steward,“Variable laterality of corticospinal tract axons that regenerateafter spinal cord injury as a result of PTEN deletion orknock-down,” The Journal of Comparative Neurology,vol. 524, no. 13, pp. 2654–2676, 2016.

[49] P. Yang, Y. Qin, W. Zhang, Z. Bian, and R.Wang, “Sensorimo-tor cortex injection of adeno-associated viral vector mediatesknockout of PTEN in neurons of the brain and spinal cordof mice,” Journal of Molecular Neuroscience, vol. 57, no. 4,pp. 470–476, 2015.

[50] Y. Liu, X. Wang, W. Li et al., “A sensitized IGF1 treatmentrestores corticospinal axon-dependent functions,” Neuron,vol. 95, no. 4, pp. 817–833.e4, 2017.

[51] J. Rodger, E. S. Drummond, M. Hellstrom, D. Robertson,and A. R. Harvey, “Long-term gene therapy causestransgene-specific changes in the morphology of regenerat-ing retinal ganglion cells,” PLoS One, vol. 7, no. 2, articlee31061, 2012.

14 Neural Plasticity

[52] C. J. LeVaillant, A. Sharma, J. Muhling et al., “Significantchanges in endogenous retinal gene expression assessed 1 yearafter a single intraocular injection of AAV-CNTF or AAV-BDNF,” Molecular Therapy Methods & Clinical Development,vol. 3, article 16078, 2016.

[53] S. A. Azizi, D. Stokes, B. J. Augelli, C. DiGirolamo, andD. J. Prockop, “Engraftment and migration of human bonemarrow stromal cells implanted in the brains of albinorats—similarities to astrocyte grafts,” Proceedings of theNational Academy of Sciences of the United States of America,vol. 95, no. 7, pp. 3908–3913, 1998.

[54] J. Sanchez-Ramos, S. Song, F. Cardozo-Pelaez et al., “Adultbone marrow stromal cells differentiate into neural cellsin vitro,” Experimental Neurology, vol. 164, no. 2, pp. 247–256, 2000.

[55] D. Woodbury, E. J. Schwarz, D. J. Prockop, and I. B. Black,“Adult rat and human bone marrow stromal cells differentiateinto neurons,” Journal of Neuroscience Research, vol. 61, no. 4,pp. 364–370, 2000.

[56] G. C. Kopen, D. J. Prockop, and D. G. Phinney, “Marrowstromal cells migrate throughout forebrain and cerebellum,and they differentiate into astrocytes after injection intoneonatal mouse brains,” Proceedings of the National Academyof Sciences of the United States of America, vol. 96, no. 19,pp. 10711–10716, 1999.

[57] W. Tetzlaff, E. B. Okon, S. Karimi-Abdolrezaee et al., “Asystematic review of cellular transplantation therapies forspinal cord injury,” Journal of Neurotrauma, vol. 28, no. 8,pp. 1611–1682, 2011.

[58] R. S. Oliveri, S. Bello, and F. Biering-Sorensen, “Mesenchy-mal stem cells improve locomotor recovery in traumaticspinal cord injury: systematic review with meta-analyses ofrat models,” Neurobiology of Disease, vol. 62, pp. 338–353,2013.

[59] M. Ye, S. Chen, X. Wang et al., “Glial cell line-derivedneurotrophic factor in bone marrow stromal cells of rat,”Neuroreport, vol. 16, no. 6, pp. 581–584, 2005.

[60] R. D. Nandoe, A. Hurtado, A. D. Levi, A. Grotenhuis, andM. Oudega, “Bone marrow stromal cells for repair of the spinalcord: towards clinical application,” Cell Transplantation,vol. 15, no. 7, pp. 563–577, 2006.

[61] C. M. Kolf, E. Cho, and R. S. Tuan, “Mesenchymal stromalcells. Biology of adult mesenchymal stem cells: regulation ofniche, self-renewal and differentiation,” Arthritis Research &Therapy, vol. 9, no. 1, p. 204, 2007.

[62] S. Hodgetts, G. W. Plant, and A. Harvey, “Chapter 14. Spinalcord injury: experimental animal models and relation tohuman therapy,” in The Spinal Cord, H. Watson, G. Paxinos,and G. Kayalioglu, Eds., Elsevier, London, 2009.

[63] A. S. Kramer, A. R. Harvey, G. W. Plant, and S. I. Hodgetts,“Systematic review of induced pluripotent stem cell technologyas a potential clinical therapy for spinal cord injury,” CellTransplantation, vol. 22, no. 4, pp. 571–617, 2013.

[64] A. Nagai, W. K. Kim, H. J. Lee et al., “Multilineage potential ofstable human mesenchymal stem cell line derived from fetalmarrow,” PLoS One, vol. 2, no. 12, article e1272, 2007.

[65] P. Lu, G. Woodruff, Y. Wang et al., “Long-distance axonalgrowth from human induced pluripotent stem cells after spinalcord injury,” Neuron, vol. 83, no. 4, pp. 789–796, 2014.

[66] S. Gronthos, A. C. W. Zannettino, S. J. Hay et al., “Molecularand cellular characterisation of highly purified stromal stemcells derived from human bone marrow,” Journal of CellScience, vol. 116, no. 9, pp. 1827–1835, 2003.

[67] S. V. White, C. E. Czisch, M. H. Han, C. D. Plant, A. R. Harvey,and G. W. Plant, “Intravenous transplantation of mesenchy-mal progenitors distribute solely to the lungs and improveoutcomes in cervical spinal cord injury,” Stem Cells, vol. 34,no. 7, pp. 1812–1825, 2016.

15Neural Plasticity

Hindawiwww.hindawi.com Volume 2018

Research and TreatmentAutismDepression Research

and TreatmentHindawiwww.hindawi.com Volume 2018

Neurology Research International

Hindawiwww.hindawi.com Volume 2018

Alzheimer’s DiseaseHindawiwww.hindawi.com Volume 2018

International Journal of

Hindawiwww.hindawi.com Volume 2018

BioMed Research International

Hindawiwww.hindawi.com Volume 2018

Research and TreatmentSchizophrenia

Hindawi Publishing Corporation http://www.hindawi.com Volume 2013Hindawiwww.hindawi.com

The Scientific World Journal

Volume 2018Hindawiwww.hindawi.com Volume 2018

Neural PlasticityScienti�caHindawiwww.hindawi.com Volume 2018

Hindawiwww.hindawi.com Volume 2018

Parkinson’s Disease

Sleep DisordersHindawiwww.hindawi.com Volume 2018

Hindawiwww.hindawi.com Volume 2018

Neuroscience Journal

MedicineAdvances in

Hindawiwww.hindawi.com Volume 2018

Hindawiwww.hindawi.com Volume 2018

Psychiatry Journal

Hindawiwww.hindawi.com Volume 2018

Computational and Mathematical Methods in Medicine

Multiple Sclerosis InternationalHindawiwww.hindawi.com Volume 2018

StrokeResearch and TreatmentHindawiwww.hindawi.com Volume 2018

Hindawiwww.hindawi.com Volume 2018

Behavioural Neurology

Hindawiwww.hindawi.com Volume 2018

Case Reports in Neurological Medicine

Submit your manuscripts atwww.hindawi.com